Development of a Deep Learning for an Inverse Design of Phase Separation Structure in Polymer Alloy

By using some machine learning methods, in this study, we develop a framework that deals with forward analysis to predict a property from a polymer alloy’s phase separation structure and inverse design to generate the structure from the property. We consider only Young’s modulus as the property in this study. In our framework, the forward analysis is performed by using a convolutional neural network (CNN) and the inverse design is realized by a random search toward a model combining a generative adversarial network (GAN) and a CNN. This framework is applicable to various properties at a low computational cost, and the latent variables belonging to the GAN are useful for feature extraction.